Recent health communications research using tailored messages, computer applications and interactive telephone outreach indicates that these approaches can effectively influence health-related behaviors, but very little research has attempted to apply these innovations to colorectal cancer (CRC) screening. Virtually no research of this type has been reported in minority populations. Screening for CRC has been shown to reduce CRC mortality, and current NCI guidelines recommend regular CRC screening for people aged 50 to 80 years. Yet the prevalence of CRC screening remains low, particularly in low-income and minority populations. The goal of the proposed study is to evaluate tailored health communications as a strategy for promoting CRC screening in a predominantly low-income minority population. The intervention, which will be delivered in Spanish as well as English, will be directed at a predominantly black and Hispanic population of men and women between 50-80 years of age who have not had CRC screening in at least the past two years. Participants will be sampled from the 1199 National Benefit Fund, the self-administered and self-insured health insurance and retirement benefit fund of the largest health care workers union in the U.S. The intervention comprises outreach and health communications provided through tailored telephone counseling coupled with computer-based decision support and tailored follow-up print communications. The proposed intervention strategy is a logical extension of recent research by Columbia University investigators and others, showing tailored health communications delivered by telephone to be an effective means of promoting screening behavior in low-income and minority populations. The study is designed as a randomized controlled trial with blinded ascertainment of medically documented CRC screening as the outcome. A total of 1,204 men and women will be randomized and followed to assess rates of fecal occult blood testing or flexible sigmoidoscopy within six months of randomization. This sample size will provide adequate statistical power to test the effect of the intervention separately in men and in women. Measures of mediating variables, including knowledge and beliefs, will be collected and used to formulate models predicting which subgroups are most and least likely to respond to the intervention. This proposal addresses the need for research assessing tailored communications to increase CRC screening in minority groups.